Abstract Background Vitamin B1 is a water-soluble vitamin required for human health. Thiamine diphosphate (TDP), the active form of vitamin B1, plays an essential role in biochemical pathways required for energy metabolism. TDP is analyzed routinely in the laboratory to assess the nutritional status of post-bariatric surgery patients and other individuals who are at risk of vitamin deficiency. In recent years, demand for vitamin B1 testing has increased dramatically, prompting development of a high throughput assay. Here, we present a method using protein precipitation and multiplex LC-ESI-MS/MS analysis at low pH with reversed-phase parameters that optimize chromatographic separation. The new method was validated for linearity, sensitivity, accuracy, imprecision, specificity, reproducibility, carryover, and stability. Method Whole blood samples were deproteinized using trichloroacetic acid (TCA) after the addition of stable-isotope-labeled TDP (d3-TDP) as an internal standard (IS). TDP was separated by reversed-phase chromatography on Waters XBridge™ Premier BEH C18 columns (2.1 × 100 mm, 2.5 µm) and analyzed on an Agilent StreamSelect™ LC-ESI-MS/MS platform. The StreamSelect™ system consists of four parallel Agilent Infinity Lab LC Series II instruments coupled to a single triple quadrupole mass spectrometer (6470B). TDP was found to interact with metal surfaces in the LC system, so the flow path was passivated using a deactivator solution which improved overall sensitivity and peak quality. TDP eluted in 3.15 ± 0.08 minutes; run times per column for a single stream and four streams were 9 min and 2.25 min, respectively. TDP was analyzed using positive electrospray ionization and multiple reaction monitoring mode with mass transitions of m/z 425.05 > 121.1 (quantifier) and m/z 425.05 > 304.00 (qualifier). Results The method was linear from 20 to 1000 nmol/L with an R2 > 0.99. Replicate analysis of low calibrators to evaluate assay sensitivity demonstrated accuracy within 10% and imprecision of ≤5%. Spike-recovery experiments showed an accuracy of ±15%. TDP measured in purchased, matrix-matched quality control (QC) and expired proficiency testing materials met the specified range for all samples. The intra- and inter-day assay imprecision was ≤3%. The cross-analyte interference study showed the analyte response in the IS-only blank was less than 10% of the LLOQ area. Also, the IS response at the ULOQ was ≤5% of the analyte area. Repeated injections of calibrators and QC materials across the four LC streams showed excellent parity (<2% imprecision). Re-injection reproducibility of extracted samples at 24 and 48 h was within ±2% of time zero concentration. Batch size tests indicated a high degree of robustness required to support the anticipated high testing volumes. Conclusions An accurate, specific, and high-throughput LC-ESI-MS/MS method was developed and validated to measure TDP in whole blood. Simple, quick sample preparation was employed for adaption to a high-throughput platform. StreamSelect™ using four parallel LCs reduced injection times to 2.25 min per sample and significantly minimized mass spectrometer idle time. The assay performance evaluations met our laboratory-developed test acceptance criteria for use in routine clinical laboratory analysis.
Read full abstract